Abstract

OBJECTIVE: To determine whether reminder cards in medical records enhance the effectiveness of audit with feedback in improving the care of patients taking long term benzodiazepine drugs. DESIGN: Randomised trial, practices receiving feedback only in one group and practices receiving feedback plus reminder cards in the other group. SETTING: 18 general practices in Leicestershire. SUBJECTS: Random samples of patients who had been taking a benzodiazepine anxiolytic or hypnotic drug for four weeks or longer. MAIN OUTCOME MEASURES: Entries in medical records indicating compliance with five criteria of care: assessment of suitability for withdrawal; being told about dependency; withdrawal being recommended; withdrawal or continuing medication; and a consultation with the general practitioner in the past year. Data were collected before and after feedback or feedback plus reminders. RESULTS: Of a total population of 125,846 registered with the 18 practices, 2409 (1.9%) had been taking a benzodiazepine for four weeks or longer. Of the 742 in the first samples, 543 (73.2%) were women, the mean (SD) age was 68.7 (14.9) years, and they had been taking a benzodiazepine for 10.1 (6.7) years. The number of patients whose care complied with the criteria rose after the interventions to implement change. The increase was greater in practices receiving feedback plus reminders for only two of the five criteria "told about dependency" increasing from 52 (11.1%) to 118 (25.8%) in the feedback only group, and from 27 (10.5%) to 184 (43.0%) in the feedback plus reminders group; odds ratio (OR) 1.46 (95% confidence interval (95% CI) 1.32 to 5.21); and "consulted in the past year" increasing from 434 (93.1%) to 411 (95.8%) in the feedback only group and 255 (96.6%) to 400 (99.8%) in the feedback plus reminders group, OR (95% CI) 13.5 (2.01 to 330.3). CONCLUSIONS: Reminder cards had only a limited effect and cannot be recommended for routine use. There were improvements in the care of patients of both groups of practices and further studies are indicated to determine the impact of both systematically developed criteria and reminders embedded into restructured medical records.